OR.

Currently building at IPMD, Inc.

Om Rajasekharan

University of North Carolina at Chapel Hill · Mathematics & Biostatistics

I build reliable software and data-driven products for complex real-world problems.

View ProjectsDownload Resume
GitHubLinkedIn

A transit-style diagram. Three lines — Engineering, Data and ML, and Research — originate at UNC-Chapel Hill and run through CloudSci, UNC IDEEL, FRC Robotics, PreconAI, MyLocalHealth, and the current role at IPMD. Lines share a station wherever that experience genuinely combined those disciplines. Tab through the stations below to read each one, or use the Auto / Explore control to start or stop the automatic tour.

University of North Carolina at Chapel Hill — B.S. Mathematics & Biostatistics, jump to educationCloudSci — Software Engineering Intern, Feb 2024 – Aug 2024. 82% classification accuracy across 100K+ frames.UNC Infectious Disease Epidemiology and Ecology Lab — Undergraduate Data Science Researcher, Feb 2025 – Present. Analysis turnaround cut from 2 weeks to 3 days.Autonomous Robotics Control System — Project. Top-10 global ranking across 5 competitive robots.PreconAI — Project. 99.9% uptime across 10,000+ project records.MyLocalHealth — Featured project. Calibration error cut 97% (ECE 0.31 -> 0.01).IPMD, Inc. — Full-Stack Developer Intern, July 2026 – Present. 12+ production features shipped across 3 products.

Tab or hover a station to explore

01About

How I Think About Problems

I study Mathematics and Biostatistics at UNC-Chapel Hill, and most of my work sits where those subjects meet software engineering. Statistics gives me a way to reason carefully about uncertainty and evidence; software is how I turn that reasoning into something people can actually use.

In practice that looks like building calibrated machine learning models and the production systems around them, writing reproducible data pipelines for lab research, and shipping full-stack features for healthcare products. I care about whether a system's outputs are actually trustworthy, not just whether it runs.

Studying

Mathematics & Biostatistics

Focus

Software, ML, healthcare data

Based

Chapel Hill, NC

02Experience

Where I've Worked

Roles where I owned production features, research pipelines, or systems end to end.

July 2026 – Present

Remote

IPMD, Inc.

Full-Stack Developer Intern

  • Developing full-stack features across AI-powered healthcare and XR applications using React, Next.js, TypeScript, Node.js, Python, and Supabase, contributing to 12+ production features across 3 products.
  • Engineered OpenCV-based facial emotion-recognition pipelines for EchoAI and Emotion Sphere, achieving 91%+ emotion-classification accuracy while processing live video streams at 24+ FPS during internal testing.
  • Built and optimized REST APIs and backend inference workflows supporting 5,000+ daily requests, reducing end-to-end prediction latency by 32% through query optimization, asynchronous processing, and caching.
ReactNext.jsTypeScriptNode.jsPythonSupabaseOpenCV

Feb 2025 – Present

Chapel Hill, NC

UNC Infectious Disease Epidemiology and Ecology Lab

Undergraduate Data Science Researcher

  • Processed and analyzed 5,000+ dried blood spot samples for a dengue seroprevalence study using Luminex Magpix immunoassay technology to quantify antibody responses across multiple antigens.
  • Developed scalable, reproducible R pipelines using tidyverse and ggplot2 to clean, analyze, and visualize 100,000+ fluorescence measurements, cutting analysis time from 2 weeks to 3 days.
Rtidyverseggplot2Luminex Magpix

Feb 2024 – Aug 2024

Fort Collins, CO

CloudSci

Software Engineering Intern

  • Designed an OpenCV/NumPy computer vision pipeline with image preprocessing and feature extraction to classify pollen particles from real-time video streams, achieving 82% accuracy across 100,000+ frames.
  • Optimized image-processing algorithms through vectorization and parallelization, reducing frame latency by 45%.
PythonOpenCVNumPy

03Projects

Featured Work

MyLocalHealth is the project I've put the most engineering depth into, from calibrated ML to a WebAssembly cross-check on production scoring.

Primary project

MyLocalHealth

A ZIP-code-level public health forecasting platform: environmental, respiratory, and community data distilled into a plain-language daily snapshot.

External data sources
7+
Calibrated ML models in production
9
Calibration error reduction (ECE)
0.31 -> 0.01
Next.jsTypeScriptPythonscikit-learnXGBoostC++/WebAssemblySupabaseSQL
View project breakdownLive siteGitHub
Raw output · ECE 0.31Platt-calibrated · ECE 0.01

PreconAI

AI-assisted construction estimator and bid matcher, built end to end as founding engineer.

Uptime across AWS deployment
99.9%
Project records managed
10,000+
ReactNode.jsMongoDBPythonAWSGoogle Cloud Storage
Live site

Autonomous Robotics Control System

Closed-loop swerve-drive autonomy for FRC Team 4499: gyro/encoder odometry and an onboard vision coprocessor driving repeatable scoring routines in the 15-second, driver-independent autonomous period.

Competitive robots programmed
5
Global ranking
Top 10
JavaPythonOpenCVJeVoisWPILib

04Research

Quantitative Research, Not Just Coursework

Rigorous, reproducible work: statistical modeling, large datasets, and experimental process.

UNC Infectious Disease Epidemiology and Ecology Lab

Undergraduate Data Science Researcher · Feb 2025 – Present · Chapel Hill, NC

Quantitative work supporting a dengue seroprevalence study, spanning lab assay processing and reproducible statistical pipelines.

  • Processed and analyzed 5,000+ dried blood spot samples using Luminex Magpix immunoassay technology to quantify antibody responses across multiple antigens.
  • Developed scalable, reproducible R pipelines using tidyverse and ggplot2 to clean, analyze, and visualize 100,000+ fluorescence measurements.
  • Cut analysis turnaround from 2 weeks to 3 days by replacing ad hoc processing with a repeatable pipeline.
Rtidyverseggplot2Luminex Magpix

Publication

LBIC Imaging of Solar Cells: Introduction to Scanning Probe-Based Imaging

Journal of Chemical Education, 100(2), 1011-1016 · Jan 2023

Developed a Python automation pipeline for LBIC microscopy data acquisition and analysis, reducing manual processing time by 80% across 500+ experimental trials.

10.1021/acs.jchemed.2c00623

05Skills

What I Work With

Grouped by area. Underlined items link to a project above where I used them.

Languages

  • Python
  • Java
  • JavaScript
  • TypeScript
  • PHP
  • R
  • SQL
  • C/C++
  • Kotlin
  • HTML/CSS
  • SAS

Frontend

  • React
  • Next.js
  • HTML/CSS

Backend & Databases

  • Node.js
  • Express
  • MongoDB
  • Supabase

Machine Learning & Data

  • TensorFlow
  • scikit-learn
  • NumPy
  • Pandas
  • OpenCV

Infrastructure & Deployment

  • AWS (EC2 / S3 / Lambda)
  • GCP
  • Vercel
  • Docker

Developer & Research Tools

  • Git
  • Tableau
  • Postman
  • Jupyter

06Education

Coursework and Honors

University of North Carolina at Chapel Hill

B.S. Mathematics & Biostatistics · Minor in Biology

Expected May 2028 · Chapel Hill, NC

Relevant coursework

  • Data Structures and Management
  • Statistical Computing and Data Management
  • Principles of Experimental Analysis
  • Discrete Mathematics
  • Linear Algebra
  • Differential Equations
  • Multivariable Calculus

Honors

Honors CarolinaDean's ListVeralto ScholarshipAssured Admission to Gillings Biostatistics Program